Geological Statistics Analysis of Population Distribution at Township Level in Henan Province, China
Total Page：16
File Type：pdf, Size：1020Kb
International Proceedings of Chemical, Biological and Environmental Engineering, Vol. 91 (2016) DOI: 10.7763/IPCBEE. 2016. V91. 10 Geological Statistics Analysis of Population Distribution at Township Level in Henan Province, China Haixia Zhang, Wei Qu , Shuwen Niu, Jinghui Qi, Liqiong Ye, Guimei Zhang The College of Earth and Environment Sciences, Lanzhou University, Lanzhou 730000, China Abstract. Based on the sixth population census data at township level, this article analyzes the population distribution of Henan province, China by the geological statistics method. The result shows that population distribution of Henan province could be divided into three types: low density in mountain areas, medium density in plain areas, and high density in urban regions. The variation functions have similar trends in the four directions of E-W, N-S, NE-SW, and NW-SE. When the distance is over 80km, the anisotropy enhances. The exponential model has the best fitting effect for the variation function. The interpolation results represent the gradient change process of population density intuitively. Terrain condition is the basic factor influencing on the population spatial pattern. High population density in urban regions are the outcomes of mutual effects between the superior geographical condition and socioeconomic development. Keywords: population distribution, township level, geological statistics, variation function, Henan Province. 1. Introduction Population distribution is a reflection of the human-earth relationship in a special space-time background. Understanding the population distribution and what determines this distribution is fundamental to understanding the relationships between humans and the environment [1]. With the advancement of modern space technology and geographic information processing technology, the study on Chinese population distribution has experienced from qualitative analysis and simple quantitative to spatial-temporal modeling [2]-[4]. Many research analyzed the population distribution at the scope of provincial level in China [5], [6]. Cities, as populated densely areas, their population distribution has been getting more attention [7], [8]. Other some researches are dedicated to explaining the factors influencing population distribution [9], [10]. Spatial scale is a basic parameter to measure population density. In the large spatial scale, we can only get the macro pattern of population distribution. The characteristics of population distribution will be represented more details in the small spatial scale. However, it is meaningless if the spatial scale is too small, for the spatial variation in population density will disappear. Most of the current researches take county-level region as the basic unit to study population distribution on national or provincial scope [2], [6], [11]. There are also some works to discuss this issue in 1km×1km grid [4], [12], [13]. Due to the dense population in urban areas, most of the related research take the street, block or building as analytic unit [14]-[16]. Based on the daily living space of the major urban-rural people, it is appropriate to take township and block as the basic unit to estimate population density. Population is dense in Henan province where the human-earth relationship is in the status of tension. Based on the method of geological statistics and the sixth population census data, the article analyzes the features and trends of population distribution at township level in Henan province and reveal what factors determines this distribution. Corresponding author. Tel.: +86-18919847995; fax: +86-8914027 E-mail address: [email protected]; [email protected]; [email protected] 63 2. Overview of Research Region Henan province is located in the middle and lower reaches of the Yellow River region. Most of the regions belong to the North China Plain. The mountains lie around its western, southern, and northern edge regions. Nanyang Basin sites in southwestern mountains. Here belongs to continental monsoon climate, with good match of light, heat, water and land resources, which provides favorable conditions for people's survival and development. Hence, here becomes the important birthplace of the Chinese agricultural civilization. Historically, Henan province has been the population gravity center of China and played an important role in ethnic integration, people movement and cultural exchange [17]. Now this province includes 17 prefecture-level cities and 159 counties (including counties, the cities and districts of county level). Until the sixth census, there was 94.03 million population, it is a densely populated province in China. Its total area was approximately 16.7×104 km2 and population density was 568 people/km2 in 2010, which is approximately 4 times higher than national average level. In 2010, the demographic urbanization level of Henan province was 38.8%, and per capita GDP in current rate was 24446 yuan [18]. Its economic development level ranks the medium in the whole country. 3. Research Method 3.1. Data source and processing Population data used in this article are derived from the sixth national census. The vector shape file in GIS format for census tracts is available from 2013 edition of the "Atlas of Henan Province" [19]. Detail methods are described as follows: 1)Townships in rural areas and municipal districts have clearly defined boundaries, we take these townships as separate units. And so do some streets in municipal districts. 2)Part of the city streets interlace with each other, and it is difficult to dissect their boundaries. We merge these streets of which boundaries are intertwined into one analytic unit. 3)Some industrial parks, farms, forest farms, et al. have separate population data, but they lack clear boundaries. In this case, we merge them into the nearest units. 4)Lastly, we get 1955 analytic units. 3.2. Geostatistics method in brief With regionalized variable as its theory base, Geostatistics is a kind of mathematical geological method using variation functions and spatial interpolation as its essential tools, researching those spatial phenomena with structured and stochastic characters [20]. Under the hypothesis of meeting the second order stationarity, Z (x) is a regionalized variable, Z(xi) and Z (xi+h) are respectively the attribute value in spatial position of xi and (xi + h), and the variation function r(h) is defined as below. N (h) 1 2 r(h) [Z(xi ) Z(xi h)] [i 1, 2, ... , N(h)] (1) 2N(h) i1 In the equation (1), h is the lag distance, which stands for the separation between samples in both distance and direction. N(h) is the number of paired comparisons at lag h. Sill, Nugget, Range and Fractal Dimension are the four basic parameters of a variation function. The Sill represents the largest variation of system properties. The Nugget represents the variability and measuring errors of variables when h is less then the minimum sampling scale. Spatial autocorrelation of regionalized variable is reflected by the Range. When h exceeds the Range, the spatial autocorrelation will disappear. The Fractal Dimension stands for the curvature of the variation function, higher the Fractal Dimension, stronger the degree of spatial autocorrelation. As the theoretical model of variogram is unknown, it need to be estimated based on effective spatial samples. Spherical, Exponential, Gaussian and Hole-effective models and so on, are the common variation function models. Based on these models, spatial interpolation can achieve the continuous spatial distribution of population. 64 4. Results and Analysis 4.1. Population density in different spatial scales It is a usual phenomenon that population is unevenly distributed in space. Larger the spatial scale, smaller the extreme ratio (the ratio of the maximum and the minimum) of attribute values and stronger the average trend. Taking the 17 prefecture-level cities as basic spatial units, the most dense city is Zhengzhou, and the most sparse one is Sanmenxia, their extreme ratio is 5.4. Taking the counties as basic spatial units, the highest population density county is Weidu District of Xuchang city, and the lowest one is Lushi county, their extreme ratio is 83.5. Taking the townships as basic spatial units, the most dense unit is 83.5 times of the most sparse one. This indicates that spatial scale shrinking can effectively decrease the homogenizing trend in population density and more accurately show the actual distribution status. As shown in Table 1. Table 1: The extreme ratios on three different space scales Population density （people/km2） spatial scale Max Max Min Min prefecture-level city 1145 213 5.4 County level 7434 88 84.5 township level 17590 21 837.6 Based on the cumulative percentage of population and land area of the 1955 township level units, we draw a Lorenz curve (Fig. 1). It displays the imbalanced phenomenon of population distribution intuitively. In the bottom left corner of the Fig. 1, about 10% of population reside in 35.16% of total area, where population is sparse. These units mainly distribute in the western, southern, and northern mountain regions of Henan province. In the upper right corner of Fig. 1, about 36.20% of population occupy 10% of area, where population is dense. These units scatter across all over province, they mainly are the urban land areas and some organizational system towns with superior geographical position. The middle part of the Lorenz curve represents the plain and basin units where agricultural production is in dominant. Fig. 1.